Dear Guangping,

The choice of the number of k-points in the "z" direction will only
matter for the the "diagon" part of a transiesta run. The electrodes
matrix elements of the density matrix (corresponding to the
TS.NumUsedAtomsLeft and TS.NumUsedAtomsRight atoms) will be frozen to
the values converged in this initial part of the run.
Once the code enters the transiesta part, or in the case of a tbtrans
calculation, the k-point sampling for the 2-dimensional Brillouin zone
will be determined by the (1,1);(1,2);(2,1);(2,2) elements of the
Monkhorst_Pack block (and their displacements).
So, in your example, the sampling would be determined by:

10   0    0.0
0   10    0.0

and all the other numbers would not be considered.

Best regards,

Frederico.

On 04/11/2010 01:50 AM, Guangping Zhang wrote:
> Dear transiesta user and developers:
> In transiesta and post-process tool tbtrans fdf files,the block about K
> points
> %block kgrid_Monkhorst_Pack 
>   10   0   0    0.0          
>   0   10   0    0.0          
>   0   0    1    0.0          
> %endblock kgrid_Monkhorst_Pack
> the third row must be 1 like in smeagol,since the system is non-periodic
> along the z transport direction??
> Thanks in advance !
> Guangping Zhang

Responder a